My UoA Commute Story
Anya Prakash
Introduction
This report tells the story of my commute to and from UoA through data collected on Google Forms after making a trip.
Firstly, which mode of transportation is best?
The density plot above shows the duration of my trips based on whether I used the bus or my car.
It is evident that driving generally takes less time than the bus, with my journeys frequently taking roughly 15 minutes in the car, as opposed to over 25 minutes on the bus.
The larger spread of the ‘Bus’ data also shows that the bus is less predictable – there is more variability in how long my commute will take when I use the bus compared to the car.
This aligns with the expectations I had going into the data collection phase. However, it is more valuable for me to analyse patterns with my travels on the bus, as I tend to take the bus on weekdays to avoid having to pay for parking.
So when in the day is the bus busiest?
Each time I recorded a data point, I considered how busy the bus was on a scale from 1 (empty) to 5 (completely full). The line graph above visualises the average crowded-ness of the bus based on the time of day I began my journey.
We can see that the bus tends to be busier from roughly 8am-9am (8-9 on the x-axis) and 3pm-5pm (15-17 on the x-axis). This makes sense, as these are the times that people are commuting to and from work/school. I also take a quite ‘central’ bus route that passes multiple high schools and intermediate schools, as well as going to Britomart Station, leading to a higher density of people on the bus during peak hours.
Also note that there is no Average Crowdedness that exceeds a score of 4. This indicates that, based on this dataset, I have little reason to expect to have to stand on the bus.
How does the day of week contribute to my journey?
Knowing which times of day are busiest is useful but next, we look at whether the day of week impacts this.
The plot above shows a heat-map of the crowdedness of the bus on average based on both time of day and day of week. The yellower the square point, the emptier the bus. We can see that the late-night bus trips tend to be the emptiest, which is to be expected.
The size of the square points indicates how many data points were used to calculate the displayed average crowdedness. I.e. a large square with ‘6’ in the centre is indicative of 6 points being aggregated to create this plot point.
I felt this was useful to note because my dataset is quite small (less than 60 data points). Thus, when using this visual story to inform my future commute decisions, it is meaningful to know whether the data visualised is the mean of multiple trips or potentially a one-off. Points that are bigger are more reliable in my eyes.
We can see that there are three day-time combinations that are most common (on Monday morning, Tuesday morning, and Thursday morning). This is explained by the fact that these are the days I have classes and must get to UoA at a particular time. Because I work on Fridays, there are no data points for this day (except one at midnight which is likely a result of me leaving UoA very late on Thursday night).
A noteworthy point with regard to this plot is that not only do I have a very small dataset, but there are also gaps present. For example, we cannot say whether or not Tuesday 4pm is a busy time to leave UoA because I have no data on this. In general, these things make it difficult to use this data to make well-informed commute decisions. Instead, it can be used to refine my travel with consideration to my current schedule – although there is no data for Friday, I do not currently need to know when is best to take the bus to/from UoA on Fridays so this lack of knowledge will not negatively impact my travel.
Which bus route is busiest?
The heat-map above shows the average duration of each of my possible bus route, split by the direction of travel. Now that I have a better sense of roughly when I should take the bus each day, this allows me to optimise my travel even further by opting to take the bus route that tends to be the quickest at roughly the time that route 27 tends to be the least crowded on the particular day.
I suspect that the 27H appears to be the busiest on the ‘Home to UoA’ direction and the 27W on the ‘UoA to Home’ direction because these are the buses I take the most frequently. I tend to prefer the 27H in the morning, because this route’s bus stop is slightly closer to my house, and I find the 27W to come more frequently when I am returning home from UoA.
Another reason that the 27T is in the middle on both directions could be that this route is shorter and, therefore, less frequent and caters to less people. So, I have fewer data points on this route, and would expect it not to be the busiest route.
Overall,
I now know that the car is faster than the bus in general BUT when I do need to take the bus to/from UoA (which is most of the time), I should try to leave UoA later in the evening to get a less crowded bus. I also know that I can almost always expect to get a seat on the bus, even during peak hours.